1. Field of the Invention
The present invention generally relates to graphics processing and more specifically to distributing and accessing vertex data.
2. Description of the Related Art
In a conventional graphics processing unit (GPU), a data assembler unit constructs batches of primitives and vertex attributes that are output to a graphics pipeline for processing. When multiple processing units are used in a graphics pipeline to perform parallel processing of the primitives, the primitives and vertex attributes are transported for longer distances through bus interconnects to reach the multiple processing units. As the number of vertex attributes increases to produce more complex and realistic images, the width or speed of the bus interconnects also increases. Additional circuitry may be needed to support the wider and/or faster transfers.
Accordingly, what is needed in the art is a technique for providing primitives and vertex attributes to the graphics pipeline while minimizing the amount of hardware and wiring required to do so.